SCIENTIFIC ABSTRACT Spectral studies of the interstellar medium (ISM) in the galaxies in the Local Group, will enable us to probe the physical conditions and composition of the ISM in and around star forming regions. Fine-structure lines of several species in the infrared give us a unique opportunity to determine temperature, density, chemical composition and energy balance of the gas, while solid state features and transitions of molecular hydrogen directly measure the amount of dust and state of the molecular gas in the vicinity of the HII regions. The strength of the major cooling lines of [CII] and [OI] will constrain the different models for photo dissociation; especially the [CII] emission arising from photo-dissociated gas at the molecular cloud surfaces is very useful for probing the warm, dense molecular cloud surfaces around star forming regions. Away from star forming regions, the [CII] line emission can be used to probe the energy balance in the cooler ISM. Solid state emission features of dust (due to PAHs, such as the 11.3 mu feature), and the quadrupole transitions of radiatively excited, hot H_2, will be used to measure the dust properties and physical state of the molecular gas surrounding the star forming regions. These, combined with the information from the cooling lines of [CII] and [OI], and the fine structure lines from the HII regions themselves will enable us to build self consistent models describing the physical conditions in and around star forming regions, including the dissociation and heating of the molecular clouds and dust. To get a large range in physical properties of the surrounding molecular material we will select both known HII regions and a few embedded star formation regions found in the IRAS survey of the Magellanic Clouds. To obtain information about the cool molecular gas observations of the 17 mu line of H_2 are planned in a companion SOT proposal by R. Laureijs et al. The 17 mu line of cool H_2 will be faint and will be tried in two specific lines of sight. Photo dissociation region (PDR) models for the LMC HII regions, indicate that the line may be stronger than in Galactic HII regions because of the lower dust content and hence the greater penetration of the UV radiation field into the molecular clouds. A search for other molecular lines will be made along the same line of sight. The selected objects are the brightest HII regions in the LMC, SMC and M33. The LMC and SMC regions have been selected in coordination with members of the other ISO consortia (ISOCAM, ISOPHOT, LWS ans SOT) to ensure maximum overlap with proposed observations for the other ISO instruments. The regions in the Magellanic Clouds fall in three categories: diffuse, extended HII regions such as 30 Doradus, compact and high excitation regions such as N88 and N81 and a couple of hot and cool molecular clouds (LIRS 36, LIRS 49 and SMC-B1#1). The seven HII regions in M33 have been chosen to ensure a wide range in galactocentric radius and hence metallicity and physical conditions. OBSERVATION SUMMARY The main instruments used for this spectroscopic study of extragalactic HII regions and the ISM in nearby galaxies will be the SWS, the LWS and ISOPHOT-S. The SWS and LWS are to be used in grating scanning mode. The total observing time for this program is around 40 hours. The first observations will be complete grating scans of 30 Doradus and NGC 604. Quick reduction of these spectra will show whether the choice of lines and estimated line strengths of the proposed study are correct. These results will determine the detailed observing strategy. The selected lines for the other HII region study are given below (table 3) and will be observed in individual grating settings in the case of the SWS. The lines are spread out in such a way over the short wavelength and over the long wavelength section of the SWS that most lines can be observed simultaneously thus improving the observing efficience greatly. The LWS will be used in full grating scan mode as this is not less efficient than having about 6 separate grating settings. Coverage of the HII regions in the Magellanic Clouds with LWS is provided by a complementary LWS proposal by P. Cox et al. At each position not covered by other proposals we will take an ISOPHOT-S exposure to get a low resolution spectrum between 2.5 and 12 mu for measuring broad spectral features such as those from PAH's. The efficiency of ISOPHOT-S is such that a very short exposure of 2 minutes provides sufficient signal to noise. Similarly we will take short PHOT exposures at 25, 60, 100 and 160 mu (PHOT-P2 in P25, PHOT-C1 in C60 and C100 and PHOT-C2 in C160) and a short CAM exposure on those positions not covered by other proposals. Exposures of 2 - 4 minutes each are sufficient and add little extra to the total observing time for this program. These will provide information about the distribution of emission in the field of view and about the IR spectrum (i.e. dust temperature) of the HII regions under study.